During the taxed process of the lighter design may be given different circumstances that cause various types of defects on the plastic surface of the lighter or on the design printed on it. Detect these defects as soon as possible and eliminate defective pieces in the production chain is of vital importance for the company. The project objective is the development of a prototype inspection system based on artificial vision for the detection of defects on the surface of lighters and taxed made on these surface.
The system must be able to detect defects in the taxed, regardless of the design and the number of different color inks to form the design. It also will provide objective criteria for the inspection and permit the exhaustive inspection of the entire factory production. The high volume of production means that it has very little time for the inspection of each piece, so it will be necessary to design algorithms that are able to detect the different kind of defects at the specified time.
This change in product quality control will allow the detection of defects before the assembled of lighters, so as to avoid the cost of assembling and filling lighters must eventually be discarded.